ARTICLE OPEN ACCESS

Imaging markers of small vessel disease andbrain frailty and outcomes in acute strokeJason P Appleton MRCP (UK) Lisa J Woodhouse MSc Alessandro Adami MD Jennifer L Becker MD FRCR

Eivind Berge MD PhD Lesley A Cala MD FRCR Ana M Casado MD Valeria Caso MD PhD

Hanne K Christensen MD PhD Robert A Dineen PhD FRCR John Gommans FRACP

Panos Koumellis MRCP FRCR Szabolcs Szatmari PhD Nikola Sprigg MD FRCP (UK)

Philip M Bath FMedSci DSc and Joanna M Wardlaw FRCR FMedSci for the ENOS Investigators

Neurologyreg 202094e439-e452 doi101212WNL0000000000008881

Correspondence

Dr Bath

Philipbath

nottinghamacuk

AbstractObjectiveTo assess the association of baseline imaging markers of cerebral small vessel disease (SVD)and brain frailty with clinical outcome after acute stroke in the Efficacy of Nitric Oxide in Stroke(ENOS) trial

MethodsENOS randomized 4011 patients with acute stroke (lt48 hours of onset) to transdermalglyceryl trinitrate (GTN) or no GTN for 7 days The primary outcome was functional outcome(modified Rankin Scale [mRS] score) at day 90 Cognition was assessed via telephone at day90 Stroke syndrome was classified with the Oxfordshire Community Stroke Project classifi-cation Brain imaging was adjudicated masked to clinical information and treatment andassessed SVD (leukoaraiosis old lacunar infarctslacunes atrophy) and brain frailty (leu-koaraiosis atrophy old vascular lesionsinfarcts) Analyses used ordinal logistic regressionadjusted for prognostic variables

ResultsIn all participants and those with lacunar syndrome (LACS 1397 348) baseline CT imagingfeatures of SVD and brain frailty were common and independently associated with unfavorableshifts in mRS score at day 90 (all participants SVD score odds ratio [OR] 115 95 confidenceinterval [CI] 107ndash124 brain frailty score OR 125 95 CI 117ndash134 those with LACS SVDscore OR 130 95 CI 115ndash147 brain frailty score OR 128 95 CI 114ndash144) Brain frailtywas associated with worse cognitive scores at 90 days in all participants and in those with LACS

ConclusionsBaseline imaging features of SVD and brain frailty were common in lacunar stroke and allstroke predicted worse prognosis after all acute stroke with a stronger effect in lacunar strokeand may aid future clinical decision-making

IdentifierISRCTN99414122

RELATED ARTICLE

EditorialBrain frailty and smallvessel disease for strokeoutcome prediction Arewe there yet

Page 191

These authors contributed equally to this work

From the Stroke Trials Unit (JPA LJW NS PMB) and Radiological Sciences Research Group (RAD) Division of Clinical Neurosciences University of Nottingham Stroke (JPANS PMB) Nottingham University Hospitals NHS Trust UK Stroke Center (AA) IRCSS Sacro CuorendashDon Calabria Hospital Negrar Verona Italy Department of Medical Imaging(JLB) College of Medicine University of Arizona Tucson Department of Internal Medicine and Cardiology (EB) Oslo University Hospital Norway School of Medicine (LAC)University ofWestern Australia Crawley Department of Neuroradiology (AMC) Division of Clinical NeurosciencesWestern General Hospital Edinburgh UK Stroke Unit (VC) SantaMaria della Misericordia Hospital University of Perugia Italy Neurology (HKC) Bispebjerg and Frederiksberg Hospital Copenhagen Denmark Department of Medicine (JG)Hawkersquos Bay District Health Board Hastings New Zealand Department of Neuroradiology (PK) Nottingham University Hospitals Queenrsquos Medical Centre UK Department ofNeurology (SS) Clinical County Emergency Hospital Targu Mures Romania and Division of Neuroimaging Sciences (JMW) Centre for Clinical Brain Sciences Dementia ResearchInstitute University of Edinburgh UK

Go to NeurologyorgN for full disclosures Funding information and disclosures deemed relevant by the authors if any are provided at the end of the article

ENOS coinvestigators are listed in the Appendix 2 at the end of the article

The Article Processing Charge was funded by the University of Nottingham

This is an open access article distributed under the terms of the Creative Commons Attribution License 40 (CC BY) which permits unrestricted use distribution and reproduction in anymedium provided the original work is properly cited

Copyright copy 2019 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology e439

Cerebral small vessel disease (SVD) is a common cause oflacunar ischemic stroke hemorrhagic stroke vascular cogni-tive impairment and dementia1 The pathophysiology of SVDdiffers from that of other stroke subtypes and is thought toreflect intrinsic damage to small perforating arterioles man-ifested as endothelial dysfunction blood-brain barrier break-down and inflammation2 Imaging markers of SVD includewhite matter hyperintensities or leukoaraiosis microbleedsprominent perivascular spaces and lacunes in addition toacute lacunar infarcts and intracerebral hemorrhage3 All arevisible on MRI while microbleeds and perivascular spaces arenot visible on CT scanning

Recently instead of considering each individual SVD fea-ture separately a summary score of SVD features was as-sociated with risk factors4 cognition5 and mobility6

Several large trials have reviewed the association betweenimaging markers of SVD and outcome7ndash9 Some analysesfocused on features seen on MRI49 while others identifiedgeneral prestroke features visible on CT (including SVDspecific) that were associated independently with pooroutcome (leukoaraiosis cerebral atrophy and old vascularlesionsinfarcts) suggesting that these might representmarkers of brain frailty7

Few trials or large observational studies have focused onoutcomes after acute lacunar stroke largely because suchpatients have better clinical outcomes compared to patientswith other ischemic stroke subtypes10ndash12 Three trials havereported outcomes after acute lacunar stroke in a total of 835participants1012 with 422 being the largest trial population ofpatients with acute lacunar stroke to date12 There are alsomodest studies using nonrandomized data from thrombolysisregisters1011 Although the Secondary Prevention of SmallSubcortical Strokes (SPS3) trial recruited 3020 patients up to180 days after symptomatic lacunar stroke it does not providedata on outcome after acute lacunar stroke13 In contrast theEfficacy of Nitric Oxide in Stroke (ENOS) trial assessed thesafety and efficacy of transdermal glyceryl trinitrate (GTN) in4011 patients with acute stroke of which a large proportion(1397 35) were of the lacunar subtype14

The aims of the present analysis were to assess the influenceof imaging markers of SVD and brain frailty at baseline onfunctional and cognitive outcome in patients with acute

stroke particularly lacunar stroke in the ENOS trial Wesought to test the following hypotheses (1) baseline imag-ing markers of SVD will be more prevalent in those withlacunar than nonlacunar stroke (2) increased baseline SVDscore will be associated with poor functional and cognitiveoutcomes after acute lacunar stroke and (3) increasedbaseline brain frailty score will be associated with poorfunctional and cognitive outcomes after acute lacunar andnonlacunar stroke

MethodsDetails on the ENOS trial protocol statistical analysis planbaseline characteristics and main results have been publishedpreviously14ndash17 In summary ENOS recruited 4011 patientsin 173 centers in 23 countries within 48 hours of stroke onsetwith high systolic blood pressure (140ndash220 mm Hg) andrandomized them to GTN 5 mg patch or no patch for 7 daysParticipants on antihypertensive medication before their in-dex event were also randomized to continue or stop thesemedications for 7 days

Standard protocol approvals registrationsand patient consentsPatients or relativescaregivers provided written consentENOS was registered (ISRCTN99414122) and approved byethics committeescompetent authorities in all participatingcountries

Study populationStroke syndrome was assessed at baseline with the Oxford-shire Community Stroke Project (OCSP) clinical classifi-cation18 We incorporated imaging findings to createa hierarchy of increasing specificity of definitions of acutelacunar stroke19 as follows patients without lacunar syn-drome (LACS n = 2614) patients with clinical LACS (n =1397) including ischemic and hemorrhagic stroke patientswith LACS with a compatible scan (n = 623) that is LACSwith an adjudicated acute lacunar infarct or if no acute in-farct visible then no alternative pathology seen to explainthe presentation and patients with LACS with corre-sponding acute lacunar infarction on imaging (n = 143)Therefore LACS with corresponding acute lacunar in-farction is a subset of LACS with compatible scan and bothare subsets of LACS

GlossaryCI = confidence interval ENOS = Efficacy of Nitric Oxide in Stroke GTN = glyceryl trinitrate IST-3 = Third InternationalStroke Trial LACI-2 = Lacunar Intervention Trial-2 LACS = lacunar syndrome MD = mean difference mRS = modifiedRankin Scale OCSP = Oxfordshire Community Stroke Project OR = odds ratio RIGHT-2 = Rapid Intervention WithGlyceryl Trinitrate in Hypertensive Stroke Trialndash2 SPS3 = Secondary Prevention of Small Subcortical Strokes SSS =Scandinavian Stroke Scale STIR = Stroke Imaging Repository SVD = small vessel disease t-MMSE = telephone Mini-MentalState Examination TICS-M = modified Telephone Interview for Cognitive Status VISTA = Virtual International Stroke TrialsArchive

e440 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

ImagingCT or MRI brain scans were performed at baseline usuallybefore randomization A second CT brain scan was performedat day 7 (end of treatment) when feasible Scans were sent tothe coordinating center and adjudicated with a set proforma7

by trained neuroradiologists or neurologists (AA JLBLAC AC RAD PK JMW) blinded to clinical detailsand randomization allocation These assessments docu-mented the location size and swelling associated with anyacute ischemic or hemorrhagic lesion and the presence ofprestroke changes including atrophy leukoaraiosis and oldvascular lesions Atrophy was assessed separately in corticaland central regions defined as 0 = absent 1 =moderate or 2 =severe and compared against a standard template7 thusproviding a maximum score of 4 Leukoaraiosis was assessedseparately in anterior and posterior brain regions20 defined as0 = no lucency 1 = lucency restricted to region adjoiningventricles or 2 = lucency covering entire region from lateralventricle to cortex providing a maximum score of 4 Oldvascular lesionsinfarcts were classified by location (egcortical striatocapsular border zone lacunar)

In addition to individual imaging markers of SVD we appliedscores adapted for CT scanning as follows SVD score com-prises 1 point each for severe leukoaraiosis (score = 2 ante-riorly andor posteriorly as above) severe atrophy (score = 2cortically andor centrally) and any old lacunar infarctslacunes (maximum 3 of 3)9 SVD score excluding atrophy(maximum 2 of 2) was also assessed because atrophy al-though related is not specific to SVD Brain frailty scorecomprises 1 point each for leukoaraiosis (score = 1 or 2 an-teriorly andor posteriorly) cerebral atrophy (score = 1 or 2cortically andor centrally) and old vascular lesionsinfarcts(maximum 3 of 3) Although there is no accepted definition ofbrain frailty we used the individual features on both CT andMRI included in the Third International Stroke Trial whichwere individually shown to be associated with poor clinicaloutcome after acute stroke7

Clinical outcomesThe primary outcome was the modified Rankin Scale (mRS)score21 a 7-level ordered categorical scale (0 = independent 6= dead) measured at day 90 Secondary outcomes at day 90included disability (Barthel Index) quality of life (health utilitystatus calculated with European Quality of Life 5-dimensions3-level scale and visual analog scale) mood (Zung DepressionScale) and cognition (telephone Mini-Mental State Examina-tion [t-MMSE]22 modified Telephone Interview for CognitiveStatus [TICS-M]23 and verbal fluency [animal naming]) Inessence t-MMSE and TICS-M assess attention and memorywhile verbal fluency assesses executive function Participantswho died by day 90 were assigned the worst score for theseoutcomes24 Safety outcomes included all-causemortality earlyneurologic deterioration (reduction of ge5 points or reductionin the consciousness domain of gt2 points from baseline to day7 on the Scandinavian Stroke Scale [SSS]) and symptomatic

hypotension Day 90 outcomes were assessed by trained blin-ded assessors via telephone at national coordinating centers

Statistical analysisData were analyzed by intention to treat16 and are given asnumber (percent) median (interquartile range) or mean(SD) Differences in baseline characteristics were assessedwith the χ2 test for categorical variables and 1-way analysis ofvariance for continuous variables

Differences between treatment group (GTN vs no GTN)effects on outcome were assessed with binary logistic re-gression multiple linear regression ordinal logistic regressionor Cox proportional hazard regression Associations betweenbaseline imaging characteristics and mRS scores and cognitiveoutcomes at day 90 were assessed with ordinal logistic re-gression and multiple linear regression respectively Statisti-cal models were adjusted for baseline prognostic covariatesincluding age sex baseline mRS score history of strokehistory of diabetes mellitus final diagnosis nitrate use base-line systolic blood pressure baseline SSS score thrombolysisfeeding status time to randomization and treatment alloca-tion (GTN vs no GTN andor continue vs stop prestrokeantihypertensives) Results are reported as odds ratio (OR) ormean difference (MD) and associated 95 confidenceintervals (CIs) or standardized regression coefficient (β) withsignificance defined as p le 005 Analyses were performed withSPSS version 24 (SPS Inc Chicago IL)

Data availabilityThe data that support the findings of this study are availablefrom the corresponding author on reasonable request Sup-plementary data are available from Dryad (tables 6ndash8 andfigures 3ndash5 doiorg105061dryad6552r2b)

ResultsA total of 1397 of 4011 (348) patients with LACS wererecruited into ENOS Baseline characteristics of LACS dif-fered from those presenting with non-LACS (table 1) com-pared to participants without LACS those with LACS wereyounger (679 [120] years) more were male (611)recruited in Asia had diabetes mellitus or were smokers andthey were less dependent at baseline with less prestroke hy-pertension TIA hyperlipidemia ischemic heart disease pe-ripheral arterial disease or atrial fibrillation Participants withLACS also had milder index events (mean SSS score 42 vs 29p lt 0001) and a longer time to randomization and fewerreceived thrombolysis than participants without LACS Mostparticipants (92) were imaged with CT At baseline therewere more acute lacunar infarctions and fewer acute hemor-rhages in those with LACS than in those with non-LACSwhile background changesmdashleukoaraiosis cerebral atrophyand old vascular lesionsmdashdid not differ between the groups(data available from Dryad table 6 doiorg105061dryad6552r2b) In participants with LACS 50 had moderate or

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e441

Table 1 Baseline clinical characteristics

All Non-LACS

LACS LACS and compatible scan LACS and acute lacunar infarct

All p Value GTN No GTN All GTN No GTN All GTN No GTN

Patients n 4011 2614 1397 695 702 623 308 315 143 71 72

Age y 703 (122) 716 (12) 679 (12) lt0001 678 (122) 681 (121) 687 (116) 682 (117) 690 (115) 668 (109) 655 (115) 680 (102)

Male n () 2297 (573) 1444 (552) 853 (611) lt0001 424 (61) 429 (611) 395 (634) 185 (601) 210 (667) 84 (587) 41 (577) 43 (597)

Geographic region n ()

Asia 559 (139) 276 (106) 283 (203) lt0001 140 (201) 143 (204) 107 (172) 57 (185) 50 (159) 46 (322) 23 (324) 23 (319)

Europe 647 (161) 409 (156) 238 (203) 025 118 (17) 120 (171) 125 (201) 62 (201) 63 (200) 17 (119) 10 (141) 7 (97)

United Kingdom 2545 (635) 1741 (666) 804 (576) lt0001 399 (574) 405 (577) 359 (576) 172 (558) 187 (594) 75 (524) 35 (493) 40 (556)

mRS score gt0 n () 1026 (256) 733 (28) 293 (21) lt0001 134 (193) 159 (226) 130 (209) 62 (201) 68 (216) 28 (196) 14 (197) 14 (194)

Medical history n ()

Hypertension 2607 (65) 1768 (676) 839 (601) lt0001 405 (583) 434 (618) 364 (584) 175 (568) 189 (600) 80 (559) 41 (577) 39 (542)

Diabetes mellitus 699 (174) 427 (163) 272 (195) 0013 134 (193) 138 (197) 137 (220) 68 (221) 69 (219) 36 (252) 22 (31) 14 (194)

Atrial fibrillation 762 (19) 626 (239) 136 (97) lt0001 63 (91) 73 (104) 54 (87) 29 (94) 25 (79) 4 (28) 3 (42) 1 (14)

Prior stroke 594 (148) 406 (155) 188 (135) 008 96 (138) 92 (131) 79 (127) 38 (123) 41 (130) 19 (133) 10 (141) 9 (125)

Prior TIA 544 (136) 387 (148) 157 (112) 0001 83 (119) 74 (105) 77 (124) 40 (130) 37 (117) 16 (112) 9 (127) 7 (97)

Prior IHD 669 (167) 467 (179) 202 (145) 0002 97 (14) 105 (15) 96 (154) 47 (153) 49 (156) 17 (119) 7 (99) 10 (139)

Prior PAD 117 (29) 79 (3) 38 (27) 0040 16 (23) 22 (31) 19 (30) 9 (29) 10 (32) 3 (21) 1 (14) 2 (28)

Hyperlipidemia 1098 (274) 759 (29) 339 (243) lt0001 173 (249) 166 (236) 155 (249) 83 (269) 72 (229) 34 (238) 16 (225) 18 (25)

Smoking current 945 (236) 566 (217) 379 (271) lt0001 186 (268) 193 (275) 181 (291) 90 (292) 91 (289) 50 (35) 23 (324) 27 (375)

Nitrate 154 (38) 115 (44) 39 (28) 0012 22 (32) 17 (24) 10 (16) 8 (26) 2 (06) 1 (07) 0 1 (14)

Alcohol gt21 unitswk 294 (73) 185 (71) 109 (78) 040 56 (81) 53 (75) 43 (69) 19 (62) 24 (76) 10 (7) 4 (56) 6 (83)

Qualifying event

Ischemic stroke n () 3342 (833) 2162 (827) 1180 (845) 016 584 (84) 596 (849) 623 (100) 308 (100) 315 (100) 143 (100) 71 (100) 72 (100)

Hemorrhagic stroke n () 629 (157) 429 (164) 200 (143) 008 100 (144) 100 (142) 0 0 0 0 0 0

SSS score (of 58) 337 (132) 294 (135) 417 (79) lt0001 419 (80) 416 (77) 429 (74) 431 (75) 428 (73) 424 (79) 429 (74) 420 (84)

NIHSS score (of 42) calculated 112 (57) 13 (58) 77 (34) lt0001 77 (34) 78 (33) 72 (32) 72 (32) 73 (32) 74 (34) 73 (32) 76 (36)

Continued

e442Neu

rology

|Vo

lume94N

umber

5|

February

42020NeurologyorgN

severe cerebral atrophy 41 had some degree of leukoar-aiosis and 61 had an old vascular lesion SVD and brainfrailty scores were moderately positively correlated (Spear-man correlation coefficient 0626 p lt 0001)

Imaging and functional outcomeThe associations between baseline imaging characteristics andmRS score at day 90 were assessed in all participants (n =3995) those with LACS (n = 1392) those with LACS anda compatible scan (n = 623) and those with LACS with anacute lacunar infarction (n = 143) (table 2) In the wholepopulation visible infarction cerebral atrophy score leukoar-aiosis score old lacunar infarctlacune and old striatocapsularinfarct were individually associated with unfavorable shifts inmRS score at day 90 In participants with LACS parenchymalhemorrhage cerebral atrophy score leukoaraiosis score oldlacunar infarctlacune and old striatocapsular infarct were in-dividually associated with unfavorable shifts in mRS score atday 90 In those with LACS and a compatible scan cerebralatrophy old lacunar infarctlacune and old striatocapsular in-farct were individually associated with unfavorable shifts inmRS score at day 90 In those with LACS and acute lacunarinfarction no individual imaging features were associated withmRS score at day 90

Increasing SVD score was associated with unfavorable shifts inmRS score at day 90 in the whole ENOS population thosewith LACS and those with LACS and a compatible scan Each1-point increase in SVD score was associated with an in-creased odds of shift in the mRS score to more death ordependency with increasing effect sizes present with in-creasing specificity of lacunar stroke diagnosis whole ENOSpopulation OR 115 (95 CI 107ndash124) LACS OR 130(95 CI 115ndash147) LACS and a compatible scan OR 143(95 CI 119ndash172) and LACS with acute lacunar infarctionOR 145 (95 CI 100ndash211) (table 2) Removing atrophyfrom the SVD score resulted in comparable associations withoutcome (data not shown) In a multivariate ordinal re-gression assessing the association between individual featurescomprising the SVD score and functional outcome in thewhole ENOS population all features had similar effect sizeswithin the model severe leukoaraiosis OR 117 (95 CI099ndash138) severe atrophy OR 112 (95 CI 099ndash127) andold lacunar infarctslacunes OR 117 (95 CI 104ndash132)However with increasing specificity of lacunar stroke di-agnosis the effect sizes for severe atrophy and old lacunarinfarctslacunes increased while severe leukoaraiosis did not(severe atrophy LACS OR 135 [95 CI 110ndash167] LACSand a compatible scan OR 139 [95 CI 102ndash189] oldlacunar infarctslacunes LACS OR 131 [95 CI 107ndash161]LACS and a compatible scan OR 167 [95 CI 123ndash225])

Increased brain frailty features were associated with worsefunctional outcome at 90 days (figure 1) in the whole ENOSpopulation (OR 125 95CI 117ndash134) in those with LACS(OR 128 95 CI 114ndash144) and in those with LACS anda compatible scan (OR 129 95 CI 108ndash153) but not inTa

ble

1Bas

elineclinical

charac

teristics(con

tinue

d)

All

Non-LACS

LACS

LACSandco

mpatible

scan

LACSandacu

telacu

narinfarct

All

pValue

GTN

NoGTN

All

GTN

NoGTN

All

GTN

NoGTN

GCSscore

lt15

n(

)12

29(306)

1145

(438)

84(6)

lt0001

48(69)

36(51)

25(40)

14(45)

11(35)

7(49)

3(42)

4(56)

Timeto

randomizationh

260

(129)

254

(131)

271

(124)

lt0001

271

(127)

271

(122)

273

(126)

278

(130)

269

(122)

322

(109)

308

(113)

336

(104)

Thro

mbolysis

n(

)42

5(106)

335(128)

90(64)

lt0001

37(53)

53(75)

46(74)

18(58)

28(89)

4(28)

3(42)

1(14)

AbbreviationsGCS=Glasg

ow

ComaSc

ale

GTN

=glyceryl

trinitrateIHD=isch

emichea

rtdisea

seL

ACS=lacu

nar

syndrome

mRS=modifiedRan

kinSc

ale

NIHSS

=NIH

Stroke

Scale

PAD=peripheral

arterial

disea

seS

SS=

Scan

dinav

ianStroke

Scale

LACSvs

non-LACSas

sessed

withχ2

forca

tego

rica

lvariablesan

d1-way

analysis

ofva

rian

ceforco

ntinuousva

riab

les

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e443

Table 2 Associations between baseline imaging characteristics and primary outcome (mRS score at day 90)

Overall LACS LACS and compatible scan LACS and acute lacunar infarct

n () OR (95 CI) p Value n () OR (95 CI) p Value n () OR (95 CI) p Value n () OR (95 CI) p Value

No 3995 mdash mdash 1392 mdash mdash 623 mdash mdash 143 mdash mdash

Visible infarction 2041(509)

114 (101 128) 0030 658(471)

117 (094 145) 015 144(241)

098 (069 138) 091 143(100)

mdash mdash

Lacunar 241 (64) 083 (066 105) 013 143(110)

083 (060 113) 024 143(240)

097 (068 137) 085 143(100)

mdash mdash

Visible hemorrhage 673(168)

110 (095 129) 020 197(141)

122 (093 161) 015 0 mdash mdash 0 mdash mdash

Parenchymal hemorrhage 587(146)

111 (095 131) 019 184(141)

138 (101 189) 0041 0 mdash mdash 0 mdash mdash

Lobar or cerebellar 79 (21) 146 (098 218) 007 25 (19) 084 (041 171) 063 0 mdash mdash 0 mdash mdash

Deep 507(134)

106 (089 125) 052 158(121)

134 (099 182) 0057 0 mdash mdash 0 mdash mdash

Cerebral atrophy score (of 4)median (IQR)

2 (2) 114 (109 121) lt0001 2 (2) 123 (113 135) lt0001 2 (2) 116 (102 132) 0028 2 (3) 112 (088 142) 083

Leukoaraiosis score (of 4) median(IQR)

0 (2) 111 (107 116) lt0001 0 (2) 111 (103 119) 0007 0 (2) 110 (098 123) 011 1 (2) 115 (091 145) 024

Old infarcts

Striatocapsular 585(152)

129 (110 151) 0002 210(158)

151 (115 197) 0003 114(183)

182 (125 264) 0002 14 (98) 169 (062 465) 031

Border zone 74 (19) 135 (089 204) 015 26 (20) 069 (035 139) 031 9 (14) 052 (016 171) 028 4 (28) 026 (004 165) 015

Lacunar 1423(370)

119 (106 135) 0003 514(386)

135 (111 165) 0003 249(400)

170 (126 229) 0001 68(476)

143 (077 264) 026

At least 1 of above 1619(421)

121 (108 136) 0001 588(441)

135 (111 165) 0003 284(456)

176 (131 236) lt0001 71(497)

131 (071 240) 038

SVD score (of 3) median (IQR) 1 (2) 115 (107 124) lt0001 1 (2) 130 (115 147) lt0001 1 (2) 143 (119 172) lt0001 1 (2) 145 (100 211) 0051

Brain frailty score (of 3) median(IQR)

2 (2) 125 (117 134) lt0001 2 (2) 128 (114 144) lt0001 2 (2) 129 (108 153) 0005 2 (2) 116 (084 160) 035

Abbreviations CI = confidence interval IQR = interquartile range LACS = lacunar syndrome mRS = modified Rankin Scale OR = odds ratio SVD = small vessel diseaseData are number (percent) or median (IQR) SVD score comprises severe anterior or posterior leukoaraiosis any old lacunar infarcts severe central or cortical atrophy (maximum 3 of 3) Brain frailty score comprisesleukoaraiosis cerebral atrophy and old vascular lesions (maximum 3 of 3) The OR represents an ordinal logistic regression analysis of shift in the mRS score with adjustment for age sex baseline Scandinavian Stroke Scalescore and time to randomization

e444Neu

rology

|Vo

lume94N

umber

5|

February

42020NeurologyorgN

the small group of LACS with an acute lacunar infarct Hencea 1-point increase in brain frailty score was associated witha similar increased odds of shift to more death or dependencyacross the 3 populations In a multivariate ordinal regressionassessing the association between individual features com-prising brain frailty and functional outcome in the whole

ENOS population all features had similar effect sizes withinthe model leukoaraiosis OR 110 (95 CI 105ndash115) atro-phy OR 112 (95 CI 106ndash119) and old vascular lesionsinfarcts OR 114 (95 CI 101ndash129) Unlike with the SVDscore we did not see any consistent changes in the effect sizesfor the individual features comprising the brain frailty score aswe progressed through the lacunar stroke hierarchy

MRI data were available for 109 participants with LACS Onlyold lacunar infarctslacunes were associated with an un-favorable shift in mRS score at day 90 (data available fromDryad table 7 doiorg105061dryad6552r2b)

Imaging and cognitive outcomesDay 90 cognitive outcomes were available in about half of theoverall population t-MMSE in 1949 (49 table 3) TICS-Min 1930 (48 table 4) and verbal fluency in 2269 (57table 5) Participants with no cognitive data at 90 days wereolder (73 [16] vs 71 [18] years) and had more severe strokes(baseline mean SSS score 35 vs 32 p lt 0001) than those withat least 1 telephone cognitive assessment

Overall visible infarction cerebral atrophy score and leu-koaraiosis score were independently associated with worsecognitive scores on t-MMSE and TICS-M at 90 days (dataavailable from Dryad figures 3 and 4 doiorg105061dryad6552r2b) In addition acute lacunar infarction parenchymaland deep hemorrhage and old lacunar infarctlacunewere eachassociated independently with worse verbal fluency scores at 90days Brain frailty was associated with worse scores on all 3cognitive measures (figure 2) Those with baseline brain frailtyscores of 3 had worse cognitive scores compared to those withno evidence of brain frailty (t-MMSE MD minus113 95 CIminus237 to 011 p = 007 TICS-M MD minus249 95 CI minus425 tominus073 p = 0006 verbal fluency MD minus172 95 CI minus297 tominus046 p = 0008) Inmultivariate linear regression assessing theassociation between individual features comprising brain frailtyand cognition leukoaraiosis and atrophy were consistentlyassociated with worse cognition across all 3 measures while oldvascular lesionsinfarcts were not In contrast to brain frailtySVD score was associated with worse verbal fluency only thosewith an SVD score of 3 had lower verbal fluency scores com-pared to those with an SVD score of 0 (MD minus212 95 CIminus366 to minus058 p = 0007 data available from Dryad figure 5doiorg105061dryad6552r2b) In multivariate analysesassessing the association between individual features of theSVD score and cognitive outcomes only severe leukoaraiosiswas associated with all 3 cognitive measures while old lacunarinfarctslacunes were associated with worse verbal fluency only

In participants with LACS leukoaraiosis and cerebral atrophyscores were associated with worse cognitive scores on all 3measures (tables 3ndash5) Brain frailty score was associated withworse cognitive outcomes across all 3 domains Those withbaseline brain frailty scores of 3 had worse cognitive scorescompared to those with no brain frailty (t-MMSE MD minus19295 CI minus389 to 006 p = 0057 TICS-M MD minus426 95 CI

Figure 1 mRS score at day 90 by brain frailty score onbaseline imaging

Boxplots of modified Rankin Scale (mRS) score at day 90 by brain frailtyscore in (A) thewhole Efficacy ofNitric Oxide in Stroke (ENOS) population (n =3995) (B) participants with lacunar syndrome (LACS n = 1392) and (C)participants with LACS with compatible scan (n = 623)

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e445

minus717 to minus136 p = 0004 verbal fluency MD minus381 95 CIminus611 to minus151 p = 0001) As in the whole ENOS populationin multivariate linear regression models the individual brainfrailty markers leukoaraiosis and atrophy were consistentlyassociated with worse cognition across all 3 measures while oldvascular lesionsinfarcts were not SVD score was associatedwith worse verbal fluency only (tables 3ndash5) which in multi-variate analyses seemed to be driven by the effect of old lacunarinfarctslacunes rather than severe leukoaraiosis or severeatrophy

In those with LACS and a compatible scan leukoaraiosis scorewas associated with worse t-MMSE and TICS-M scores while

cerebral atrophy score was associated with worse verbal flu-ency scores only Brain frailty score was associated with worset-MMSE and TICS-M scores but was not significantly asso-ciated with verbal fluency SVD scores were not associatedwith cognitive outcomes in this small population (tables 3ndash5)

There were insufficient cognitive data for analysis in thosewith LACS and an acute lacunar infarction (n = 60)

GTN and lacunar strokeGTN had no effect on mRS score at day 90 compared with noGTN in those with LACS (OR 107 95 CI 088ndash129 n =1392) LACS and compatible scan (OR 109 95CI 082ndash145

Table 3 Associations between baseline imaging characteristics and t-MMSE score at day 90

Overall LACS LACS and compatible scan

n () MD (95 CI)β p Value n () MD (95 CI)β p Value n () MD (95 CI)β p Value

No 1949 mdash mdash 719 mdash mdash 313 mdash mdash

Visible infarction 996(511)

minus075 (minus129minus022)

0006 327(473)

minus024 (minus106058)

056 39(126)

minus035 (minus223153)

071

Lacunar 81(42)

minus070 (minus206067)

032 39(56)

006 (minus184196)

095 39(126)

minus035 (minus223153)

071

Visible hemorrhage 336(171)

minus061 (minus134013)

011 90(129)

minus095 (minus220030)

064 0 mdash mdash

Parenchymalhemorrhage

296(152)

minus042 (minus120036)

029 84(122)

minus062 (minus191066)

034 0 mdash mdash

Lobar or cerebellar 52(27)

minus005 (minus167157)

095 15(22)

103 (minus185392)

048 0 mdash mdash

Deep 243(125)

minus042 (minus130045)

034 68(98)

minus108 (minus250033)

013 0 mdash mdash

Cerebral atrophy score(of 4) median (IQR)a

2 (2) minus0045 0026 2 (2) minus0094 0023 2 (1) minus0066 029

Leukoaraiosis score(of 4) median (IQR)a

0 (2) minus0085 lt0001 0 (2) minus0129 lt0001 0 (2) minus0141 0013

Old infarcts

Striatocapsular 316(161)

031 (minus042104)

041 129(186)

007 (minus101115)

090 66(211)

minus069 (minus221082)

037

Border zone 35(18)

minus072 (minus280135)

049 11(16)

minus094 (minus427239)

058 3 (10) 040 (minus552632)

089

Lacunar 764(388)

minus026 (minus081029)

036 282(406)

003 (minus084090)

095 123(393)

016 (minus116147)

082

At least 1 of above 859(437)

minus014 (minus068041)

062 323(465)

minus005 (minus090080)

091 145(463)

minus058 (minus186069)

037

SVD score (of 3)median (IQR)a

1 (2) minus0016 040 1 (2) minus0042 029 1 (1) minus0060 033

Brain frailty score (3)median (IQR)a

2 (2) minus0062 0002 2 (2) minus0096 0018 2 (2) minus0157 0013

Abbreviations CI = confidence interval IQR = interquartile range LACS = lacunar syndrome MD = mean difference SVD = small vessel disease t-MMSE =telephone Mini-Mental State ExaminationSVD score comprises severe anterior or posterior leukoaraiosis any old lacunar infarcts and severe central or cortical atrophy (maximum3 of 3) Brain frailtyscore comprises leukoaraiosis cerebral atrophy and old vascular lesions (maximum 3 of 3) Multiple linear regression with adjustment for age sex baselineScandinavian Stroke Scale score and time to randomizationa Standardized regression coefficient reported

e446 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

n = 623) or LACS and acute lacunar infarction (OR 100 95CI 053ndash187 n = 143) GTNwithin 6 hours did not change anyday 90 clinical outcomes in lacunar stroke populations (dataavailable from Dryad table 8 doiorg105061dryad6552r2b)Furthermore in 339 participants with LACS GTN did notinfluence imaging markers at day 7 (data not shown)

DiscussionIn this large population of patients with lacunar stroke syn-dromes randomized into the ENOS trial baseline imagingmarkers of SVD and brain frailty were common and associated

with poor functional and cognitive outcomes at 90 days in-dividually and in combination Themagnitude of SVD and brainfrailty was similar in those with lacunar and nonlacunar strokedespite the patients with lacunar stroke being younger Fur-thermore the strength of association of SVD score with poorfunctional outcome increased with increasing specificity of la-cunar stroke diagnosis while brain frailty had similar associa-tions across the trial population GTN did not alter functionaloutcome of patients with acute stroke presenting with LACS

Prestroke TIA ischemic heart disease peripheral arterialdisease and atrial fibrillation were all less common in partic-ipants with lacunar stroke than in those with nonlacunar

Table 4 Associations between baseline imaging characteristics and TICS-M score at day 90

Overall LACS LACS and compatible scan

n () MD (95 CI)β p Value n () MD (95 CI)β p Value n () MD (95 CI)β p Value

No 1930 mdash mdash 686 mdash mdash 308 mdash mdash

Visible infarction 987(511)

minus084 (minus160minus008)

0031 326(475)

minus023 (minus145098)

071 38 (123) 088 (minus198373)

055

Lacunar 80(41)

minus057 (minus251138)

057 38(55)

120 (minus162402)

041 38 (123) 088 (minus198373)

055

Visible hemorrhage 331(170)

minus072 (minus177033)

018 88(128)

minus055 (minus240131)

056 0 mdash mdash

Parenchymalhemorrhage

292(151)

minus035 (minus146076)

053 82(120)

minus011 (minus203180)

091 0 mdash mdash

Lobar or cerebellar 52(27)

070 (minus160299)

055 15(22)

153 (minus274580)

048 0 mdash mdash

Deep 239(124)

minus056 (minus180068)

038 66(96)

minus066 (minus278145)

054 0 mdash mdash

Cerebral atrophyscore(of 4) median (IQR)a

2 (2) minus0068 0001 2 (2) minus0137 0001 2 (1) minus0095 012

Leukoaraiosis score(of 4) median (IQR)a

0 (2) minus0088 lt0001 0 (2) minus0124 0001 0 (2) minus0131 0018

Old infarcts

Striatocapsular 310(159)

029 (minus076133)

059 126(183)

minus042 (minus209119)

061 64 (206) minus141 (minus371089)

023

Border zone 35(18)

minus079 (minus373215)

060 11(16)

minus189 (minus679301)

045 3 (10) 103 (minus785991)

082

Lacunar 758(389)

minus055 (minus133024)

017 279(404)

minus052 (minus181077)

043 121(389)

minus014 (minus213186)

089

At least 1 of above 850(436)

minus038 (minus116039)

033 319(462)

minus054 (minus180072)

040 143(460)

minus085 (minus278107)

038

SVD score (of 3)median (IQR)a

1 (2) minus0024 021 1 (2) minus0063 010 1 (1) minus0063 029

Brain frailty score (of3) median (IQR)a

2 (2) minus0080 lt0001 2 (2) minus0133 0001 2 (2) minus0187 0003

Abbreviations CI = confidence interval IQR = interquartile range LACS = lacunar syndrome MD = mean difference SSS = Scandinavian Stroke Scale SVD =small vessel disease TICS-M = modified Telephone Interview for Cognitive StatusSVD score comprises severe anterior or posterior leukoaraiosis any old lacunar infarcts and severe central or cortical atrophy (maximum 3 of 3) Brain frailtyscore comprises leukoaraiosis cerebral atrophy and old vascular lesions (maximum 3 of 3) Multiple linear regression with adjustment for age sex baselineSSS score and time to randomizationa Standardized regression coefficient reported

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e447

strokes supporting previous studies demonstrating that largeartery disease and cardioembolic sources are important riskfactors for nonlacunar but less so for lacunar ischemicstrokes192526 While smoking and diabetes mellitus weremore common in our lacunar than nonlacunar stroke pop-ulation4 perhaps surprisingly those with LACS had less hy-pertension and hyperlipidemia than those with non-LACSalthough this finding is in keeping with previous data showingthat traditional vascular risk factors when combined accountfor lt2 of the variance in SVD features in patients with strokeand healthy older populations26

In line with our results SVD scores and their component im-aging findings have been associated with adverse clinical

outcomes after stroke in 4 other smaller cohorts Data from theStroke Imaging Repository (STIR)Virtual International StrokeTrials Archive (VISTA) showed that severe leukoaraiosis andtotal SVD score on MRI in 259 patients with ischemic stroketreated with thrombolysis were associated with increased dis-ability and functional dependency at 90 days9 In contrastlacunes cerebral atrophy and enlarged perivascular spaces werenot individually associated with clinical outcome probably be-cause of a lack of power A retrospective cohort involving 1026participants using MRI data found an association between SVDscore and all-cause and stroke-related mortality27 SVD burdenonMRI has also been associated with worse quality-of-life scores3 months after acute ischemic stroke28 and decreased cognitivefunction over 4 years in patients with first-ever lacunar ischemic

Table 5 Associations between baseline imaging characteristics and verbal fluency at day 90

Overall LACS LACS and compatible scan

n () MD (95 CI)β p Value n () MD (95 CI)β p Value n () MD (95 CI)β p Value

No 2269 mdash mdash 818 mdash mdash 366 mdash mdash

Visible infarction 1176(518)

minus072 (minus129minus015)

0013 395(483)

minus079 (minus182023)

013 60(164)

minus197 (minus411017)

007

Lacunar 118(52)

minus132 (minus260minus004)

0043 60(73)

minus113 (minus309084)

026 60(164)

minus197 (minus411017)

007

Visible hemorrhage 406(177)

minus097 (minus173minus021)

0013 115(140)

minus139 (minus288010)

007 0 mdash mdash

Parenchymalhemorrhage

361(159)

minus087 (minus167minus007)

0033 108(132)

minus128 (minus281026)

010 0 mdash mdash

Lobar or cerebellar 57(25)

115 (minus062292)

020 16(20)

235 (minus151620)

023 0 mdash mdash

Deep 303(134)

minus122 (minus210minus034)

0006 91(111)

minus200 (minus365minus036)

0017 0 mdash mdash

Cerebral atrophy score(of 4) median (IQR)a

2 (2) minus0075 lt0001 2 (2) minus0136 lt0001 2 (2) minus0127 0028

Leukoaraiosis score(of 4) median (IQR)a

0 (2) minus0095 lt0001 0 (2) minus0109 0002 0 (2) minus0073 018

Old infarcts

Striatocapsular 347(152)

minus034 (minus113045)

039 139(169)

minus011 (minus149126)

087 71(192)

minus081 (minus287124)

044

Border zone 42(18)

minus149 (minus366069)

018 14(17)

minus404 (minus803minus006)

0047 5(14)

minus512 (minus1169145)

013

Lacunar 867(379)

minus095 (minus154minus037)

0001 322(391)

minus116 (minus223minus008)

0035 142(384)

minus194 (minus365minus023)

0026

At least 1 of above 968(423)

minus084 (minus142minus027)

0004 365(443)

minus102 (minus207003)

0058 165(446)

minus196 (minus362minus029)

0022

SVD score (of 3)median (IQR)a

1 (2) minus0061 0001 1 (2) minus0086 0019 1 (1) minus0105 007

Brain frailty score (of3) median (IQR)a

2 (2) minus0096 lt0001 2 (2) minus0140 lt0001 2 (2) minus0108 007

Abbreviations CI = confidence interval IQR = interquartile range LACS = lacunar syndrome MD = mean difference SSS = Scandinavian Stroke Scale SVD =small vessel diseaseData are mean (SD) SVD score comprises severe anterior or posterior leukoaraiosis any old lacunar infarcts and severe central or cortical atrophy(maximum3of 3) Brain frailty score comprises leukoaraiosis cerebral atrophy and old vascular lesions (maximum3of 3)Multiple linear regression ofMD inscores with adjustment for age sex baseline SSS score and time to randomizationa Standardized regression coefficient reported

e448 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

stroke and hypertension29 In the present analysis SVD scoreswith and without atrophy showed similar associations withfunctional outcome at 90 days overall and in the lacunar strokepopulations suggesting that the main drivers for functionaloutcome were the vascular rather than the associated neurode-generative imaging features The strength of association of SVDscore with worse functional outcome increased with increasingspecificity of lacunar stroke diagnosis

We confirm the important prognostic value of the 3 brainfrailty measures on CT of leukoaraiosis atrophy and oldvascular lesions that were each independently associated withpoor outcome in IST-37 Furthermore we have demonstratedthat pooling these imaging markers in a score including anyold infarct (not just lacunes as in the SVD score) was asso-ciated with functional and cognitive outcomes 90 days afterstroke in contrast to the SVD score brain frailty showeda similar strength of association with poor functional outcomeacross non-LACS and LACS populations Although SVD andbrain frailty scores were moderately positively correlated andmeasure similar imaging markers the differences in severity ofthe individual markers included provide 2 different ways ofassessing brain health This is supported by their differingeffects on clinical outcomes and the different contribution ofthe individual features in multivariate regression models

There were interesting differences in associations betweenimaging features and performance in the different cognitivedomains In the whole ENOS population several acute andprestroke imaging features were associated with impairment inall cognitive domains However the SVD score added littlecompared with leukoaraiosis alone which tended to be asso-ciated with verbal fluency whereas brain frailty score was as-sociated with all cognitive impairments This may reflect theknown effects of white matter lesions on processing speed andof brain atrophy (as a sign of neurodegeneration) on memoryalternatively it could reflect a lack of statistical power due tomissing cognitive data Furthermore these findings suggest thatimaging signs may have differential relationships with cognitionand its domains in different stroke subtypes this emphasizesthe importance of testing all cognitive domains in future re-search and that some cognitive tests may be insensitive to thetypes of impairment specific to particular stroke types

These imaging markers are easy to detect on plain CT byphysicians and radiologists in acute stroke and given theirstrong prognostic significance may prove useful for predictingoutcome when added to clinical markers in clinical practiceWhether brain frailty on imaging correlates with clinical frailtyis unclear although these features correlate with gait bal-ance6 and cognitive impairments5 implying that a correlationwith clinical frailty is likely thus these brain features mayprove to be useful surrogate markers For future acute strokeclinical trials minimization based on baseline imaging mark-ers of brain frailty could be important to balance theseprognostic variables between treatment groups

The favorable effect of GTN given within 6 hours on strokeonset3031 was not seen in this analysis of participants withLACS The Rapid Intervention With Glyceryl Trinitrate inHypertensive Stroke Trialndash2 (RIGHT-2 ISRCTN26986053)will provide further detail on whether the effects of GTN varybetween differing stroke etiologies with imaging markersbeing key secondary outcomes3233 In addition the longer-term administration of isosorbide mononitrate (a long-actingnitrate) is being assessed for safety and efficacy in patients

Figure 2 Cognitive scores at day 90 by brain frailty score onbaseline imaging

Boxplots of cognitive scoresmdash(A) telephone Mini-Mental State Examination(t-MMSE n = 1949) (B) modified Telephone Interview for Cognitive Status(TICS-M n = 1930) and (C) verbal fluency (n = 2269)mdashat day 90 by brainfrailty in the whole Efficacy of Nitric Oxide in Stroke (ENOS) populationLower cognitive scores indicate worse cognition

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e449

with lacunar ischemic stroke and SVD in the ongoing LacunarIntervention Trial-2 (LACI-2 ISRCTN14911850)

The strengths of this ENOS analysis include the largestdataset of patients with acute lacunar stroke to date froma high-fidelity randomized controlled trial with near-completefollow-up blinded and standardized adjudication of imagingby trained observers using a standardized proforma ordinalanalysis of the mRS score to increase statistical power andgeneralizability to clinical practice through the predominantuse of CT imaging However there are important limitationsFirst no adjustment was made for multiplicity of testingTherefore some of the results may in part be due to chancealthough the strength of associations seen mitigates the risksof multiple testing Second the mean age in ENOS was lowerthan that seen in many unselected clinical stroke populationsalthough this is typical for lacunar stroke This may haveattenuated the observed associations because brain frailty islikely to be evenmore prevalent in an older population ThirdENOS recruited over a 12-year period in which clinicalpractice changed Thus the time from baseline imaging tostroke onset was longer than we would expect in currentstroke clinical practice but is still common in patients withminor stroke Fourth MRI which is more sensitive to featuresof SVD and acute infarction was performed in only a smallproportion of ENOS participants However the predominantuse of CT enabled associations between SVD features visibleon CT and outcome to be assessed which were found to be inkeeping with MRI-based studies and immediately applicablein clinical practice Fifth clinical stroke syndrome classifica-tion with the OCSP was investigator reported and not adju-dicated centrally In addition the OCSP is known tomisclassify asymp15 of lacunar strokes as partial anterior circu-lation syndrome and cortical strokes as LACS thus addingnoise to the data We accounted for this in our more specificlacunar stroke populations this added to the generalizabilityof the dataset and its findings to clinical practice Sixth tele-phone cognition data were available for about half of partic-ipants largely because of stroke severity24 limitinggeneralizability Finally although trained neuroradiologistsadjudicated the imaging data we cannot exclude interraterand intrarater variability over the time scale of the trial

We add to the increasing body of evidence that baseline im-aging markers of SVD and brain frailty are common and as-sociated with worse functional and cognitive outcomes at 90days individually and when amalgamated as scores Whetherthe vascular or neurodegenerative features are more associatedwith cognitive impairments requires further testing CT imag-ing features of brain frailty and SVD predict prognosis shouldbe considered as components of minimization in clinical trialsand may aid clinical decision-making in the future

AcknowledgmentThe authors thank the participants and investigators who tookpart in the ENOS trial

Study fundingENOS was funded by Bupa Foundation and the MedicalResearch Council (G0501797)

DisclosureJ Appleton is funded by National Institute for Health Re-search Triple Antiplatelets for Reducing Dependency AfterIschaemic Stroke (1010424) and British Heart FoundationRIGHT-2 (CS14430972) L Woodhouse A Adami JBecker E Berge L Cala A Casado V Caso H ChristensenR Dineen J Gommans P Koumellis S Szatmari and NSprigg report no disclosures relevant to the manuscript PBath was chief investigator of Medical Research CouncilENOS is a National Institute for Health research senior in-vestigator and has funding for British Heart FoundationRIGHT-2 EU FP7 EuroHYP and EU H2020 Prevention ofComplications to Improve Outcome in Elderly Patients WithAcute Stroke trials J Wardlaw was a grant applicant on ENOSwith funding from EU H2020 SVDsTarget FoundationLeducq (CVD 1605) and UK Medical Research CouncilDementia Research Institute Go to NeurologyorgN for fulldisclosures

Publication historyReceived by Neurology December 21 2018 Accepted in final formAugust 16 2019

Appendix 1 Authors

Name Location Role Contribution

Jason PAppletonMRCP (UK)

University ofNottinghamUK

Author Performed thestatistical analysesinterpreted thedata and wrotethe first draft of themanuscript

Lisa JWoodhouseMSc

University ofNottinghamUK

Author Trial statisticianrevised themanuscript

AlessandroAdami MD

Don CalabriaHospitalVerona Italy

Author Adjudicated theimaging in theENOS trial revisedthe manuscript

Jennifer LBecker MDFRCR

University ofArizonaTucson

Author Adjudicated theimaging in theENOS trial revisedthe manuscript

Eivind BergeMD PhD

Oslo UniversityHospitalNorway

Author ENOS trial SteeringCommitteemember revisedthe manuscript

Lesley ACala MDFRCR

University ofWesternAustraliaCrawley

Author Adjudicated theimaging in theENOS trial revisedthe manuscript

Ana MCasado MD

WesternGeneralHospitalEdinburgh UK

Author Adjudicated theimaging in theENOS trial revisedthe manuscript

e450 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

Appendix 1 (continued)

Name Location Role Contribution

Valeria CasoMD PhD

University ofPerugia Italy

Author ENOS trial SteeringCommitteemember revisedthe manuscript

Hanne KChristensenMD PhD

BispebjergHospitalCopenhagenDenmark

Author ENOS trial SteeringCommitteemember revisedthe manuscript

Robert ADineen PhDFRCR

University ofNottinghamUK

Author Adjudicated theimaging in theENOS trial revisedthe manuscript

JohnGommansFRACP

Hawkersquos BayDistrict HealthBoardHastings NewZealand

Author ENOS trial SteeringCommitteemember revisedthe manuscript

PanosKoumellisMRCP FRCR

NottinghamUniversityHospitals NHSTrust UK

Author Adjudicated theimaging in theENOS trial revisedthe manuscript

SzabolcsSzatmariPhD

Clinical CountyEmergencyHospital TarguMuresRomania

Author ENOS trial SteeringCommitteemember revisedthe manuscript

NikolaSprigg MDFRCP (UK)

University ofNottinghamUK

Author ENOS trial SteeringCommitteemember revisedthe manuscript

Philip MBathFMedSci DSc

University ofNottinghamUK

Author jointsenior authorcorrespondingauthor

Conceived thestudy revised themanuscript andchief investigatorof the ENOS trial

Joanna MWardlawFRCRFMedSci

University ofEdinburgh UK

Author jointsenior author

Conceived thestudy ledadjudication of theimaging in theENOS trial revisedthe manuscript

Appendix 2 Coinvestigators

Name Location Role Contribution

D Thomas UnitedKingdom

ENOS trialSteeringCommittee

Independent chair to2006

G Venables Sheffield UK ENOS trialSteeringCommittee

Independent chairfrom 2006

P Amarenco France ENOS trialSteeringCommittee

Independentphysician

K Muir Glasgow UK ENOS trialSteeringCommittee

Independentphysician UK

Appendix 2 (continued)

Name Location Role Contribution

KR Lees Glasgow UK ENOS trialSteeringCommittee

Trial SteeringCommittee member

S Pocock London UK ENOS trialSteeringCommittee

Trial statistician from2003

A Skene UnitedKingdom

ENOS trialSteeringCommittee

Trial statistician to2003

A Shone NottinghamUK

ENOS trialsteeringcommittee

Sponsorrsquosrepresentative

D Whynes UnitedKingdom

ENOS trialSteeringCommittee

Health economist

M Beridze Georgia InternationalAdvisoryCommittee

Responsible for trialconduct in Georgia

C Bladin Australia InternationalAdvisoryCommittee

Responsible for trialconduct in Australia

C Chen Singapore InternationalAdvisoryCommittee

Responsible for trialconduct in Singapore

HM Chang Singapore InternationalAdvisoryCommittee

Responsible for trialconduct in Singapore

R Collins Ireland InternationalAdvisoryCommittee

Responsible for trialconduct in Ireland

ACzlonkowska

Poland InternationalAdvisoryCommittee

Responsible for trialconduct in Poland

E Dıez-Tejedor

Spain InternationalAdvisoryCommittee

Responsible for trialconduct in Spain

A El Etribi Egypt InternationalAdvisoryCommittee

Responsible for trialconduct in Egypt

AR Ghani Malaysia InternationalAdvisoryCommittee

Responsible for trialconduct in Malaysia

AC Laska Sweden InternationalAdvisoryCommittee

Responsible for trialconduct in Sweden

J Navarro Philippines InternationalAdvisoryCommittee

Responsible for trialconduct in thePhilippines

G Ntaios AthensGreece

InternationalAdvisoryCommittee

Responsible for trialconduct in Greece

S Ozturk Turkey InternationalAdvisoryCommittee

Responsible for trialconduct in Turkey

S Phillips Canada InternationalAdvisoryCommittee

Responsible for trialconduct in Canada

Continued

NeurologyorgN Neurology | Volume 94 Number 5 | February 4 2020 e451

References1 Wardlaw J Smith C Dichgans M Mechanisms of sporadic cerebral small vessel

disease insights from neuroimaging Lancet Neurol 201312483ndash4972 Ostergaard L Sondergaard T Moreton F et al Cerebral small vessel disease capillary

pathways to stroke and cognitive decline J Cereb Blood Flow Metab 201636302ndash325

3 Wardlaw JM Doubal FN Valdes-Hernandez M et al Blood-brain barrier perme-ability and long-term clinical and imaging outcomes in cerebral small vessel diseaseStroke 201344525ndash527

4 Staals J Makin S Doubal F Dennis M Wardlaw J Stroke subtype vascular riskfactors and total MRI brain small-vessel disease burden Neurology 2014831228ndash1234

5 Staals J Booth T Morris Z et al Total MRI load of cerebral small vessel disease andcognitive ability in older people Neurobiol Aging 2015362806ndash2811

6 Pinter D Ritchie SJ Doubal F et al Impact of small vessel disease in the brain on gaitand balance Sci Rep 2017741637

7 IST-3 Collaborative Group Association between brain imaging signs early and lateoutcomes and response to intravenous alteplase after acute ischaemic stroke in theThird International Stroke Trial (IST-3) secondary analysis of a randomised con-trolled trial Lancet Neurol 201514485ndash496

8 Sato S Delcourt C Heeley E et al Significance of cerebral small-vessel disease inacute intracerebral hemorrhage Stroke 201647701ndash707

9 Arba F Inzitari D Ali M et al Small vessel disease and clinical outcomes after IV rt-PAtreatment Acta Neurol Scand 201713672ndash77

10 Pantoni L Fierini F Poggesi A Thrombolysis in acute stroke patients with cerebralsmall vessel disease Cerebrovasc Dis 2014375ndash13

11 Eggers CCJ Bocksrucker C Seyfang L The efficacy of thrombolysis in lacunar strokeevidence from the Austrian Stroke Unit Registry Eur J Neurol 201724780ndash787

12 Sprigg N Gray LJ Bath PMW et al Stroke severity early recovery and outcome areeach related with clinical classification of stroke data from the ldquoTinzaparin in AcuteIschaemic Stroke Trialrdquo (TAIST) J Neurol Sci 200725454ndash59

13 SPS3 Investigators Benavente OR Hart RG McClure LA Szychowski JM CoffeyCS Pearce LA Effects of clopidogrel added to aspirin in patients with recent lacunarstroke N Engl J Med 2012376817ndash825

14 Bath P Woodhouse L Scutt P et al Efficacy of nitric oxide with or without con-tinuing antihypertensive treatment for management of high blood pressure in acutestroke (ENOS) a partial-factorial randomised controlled trial Lancet 2015385617ndash628

15 ENOS Trial Investigators Glyceryl trinitrate vs control and continuing vs stoppingtemporarily prior antihypertensive therapy in acute stroke rationale and design of theEfficacy of Nitric Oxide in Stroke (ENOS) trial (ISRCTN99414122) Int J Stroke20061245ndash249

16 Bath PM Houlton A Woodhouse L et al Statistical analysis plan for the ldquoEfficacy ofNitric Oxide in Strokerdquo (ENOS) trial Int J Stroke 20149372ndash374

17 ENOS Investigators Baseline characteristics of the 4011 patients recruited into theEfficacy of Nitric Oxide in Stroke (ENOS) trial Int J Stroke 20149711ndash720

18 Bamford J Sandercock P Dennis M Burn J Warlow C Classification and naturalhistory of clinically identifiable subtypes of cerebral infarction Lancet 19913371521ndash1526

19 Lindley RI Wang JJ Wong MC et al Retinal microvasculature in acute lacunarstroke a cross-sectional study Lancet Neurol 20098628ndash634

20 van Swieten JC Hijdra A Koudstaal PJ van Gijn J Grading white matter lesions onCT and MRI a simple scale J Neurol Neurosurg Psychiatry 1990531080ndash1083

21 Bruno A Shah N Lin C et al Improving modified Rankin Scale assessment witha simplified questionnaire Stroke 2011411048ndash1050

22 Roccaforte MD Burke WJ Bayer BL Wengel SP Validation of a telephone version ofthe Mini-Mental State Examination J Am Geriatr Soc 199240697ndash702

23 Desmond DW Tatemichi TK Hanzawa L The Telephone Interview for CognitiveStatus (TICS) reliability and validity in a stroke sample Int J Geriatr Psychiatry 19949803ndash807

24 Ankolekar S Renton C Sare G et al Relationship between poststroke cognitionbaseline factors and functional outcome data from ldquoEfficacy of Nitric Oxide inStrokerdquo trial J Stroke Cerebrovasc Dis 2014231821ndash1829

25 Jackson CA Hutchison A Dennis MS et al Differing risk factor profiles of ischemicstroke subtypes evidence for a distinct lacunar arteriopathy Stroke 201041624ndash629

26 Wardlaw JM Allerhand M Doubal FN et al Vascular risk factors large-artery ath-eroma and brain white matter hyperintensities Neurology 2014821331ndash1338

27 Song TJ Kim J Song D et al Total cerebral small-vessel disease score is associatedwith mortality during follow-up after acute ischemic stroke J Clin Neurol 201713187ndash195

28 Liang Y Chen YK Deng M et al Association of cerebral small vessel disease burdenand health-related quality of life after acute ischemic stroke Front Aging Neurosci20179372

29 Huijts M Duits A van Oostenbrugge RJ Kroon AA de Leeuw PW Staals J Accu-mulation of MRI markers of cerebral small vessel disease is associated with decreasedcognitive function a study in first-ever lacunar stroke and hypertensive patients FrontAging Neurosci 2013572

30 Woodhouse L Scutt P Krishnan K et al Effect of hyperacute administration (within 6hours) of transdermal glyceryl trinitrate a nitric oxide donor on outcome after strokesubgroup Analysis of the Efficacy of Nitric Oxide in Stroke (ENOS) trial Stroke 2015463194ndash3201

31 Ankolekar S FullerM Cross I et al Feasibility of an ambulance-based stroke trial andsafety of glyceryl trinitrate in ultra-acute stroke the Rapid Intervention With GlycerylTrinitrate in Hypertensive Stroke Trial (RIGHT ISRCTN66434824) Stroke 2013443120ndash3128

32 Appleton JP Scutt P Dixon M et al Ambulance-delivered transdermal glyceryltrinitrate versus sham for ultra-acute stroke rationale design and protocol for theRapid Intervention with Glyceryl trinitrate in Hypertensive stroke Trial-2 (RIGHT-2)trial (ISRCTN26986053) Int J Stroke 201914191ndash206

33 Bath PM Scutt P Anderson CS et al Prehospital transdermal glyceryl trinitrate inpatients with ultra-acute presumed stroke (RIGHT-2) an ambulance-based rando-mised sham-controlled blinded phase 3 trial Lancet 20193931009ndash1020

Appendix 2 (continued)

Name Location Role Contribution

K Prasad India InternationalAdvisoryCommittee

Responsible for trialconduct in India

HA de Silva Sri Lanka InternationalAdvisoryCommittee

Responsible for trialconduct in Sri Lanka

L Wong Hong Kong InternationalAdvisorycommittee

Responsible for trialconduct in HongKong

Y-J Wang China InternationalAdvisoryCommittee

Responsible for trialconduct in China

e452 Neurology | Volume 94 Number 5 | February 4 2020 NeurologyorgN

DOI 101212WNL0000000000008881202094e439-e452 Published Online before print December 27 2019Neurology

Jason P Appleton Lisa J Woodhouse Alessandro Adami et al Imaging markers of small vessel disease and brain frailty and outcomes in acute stroke

This information is current as of December 27 2019

ServicesUpdated Information amp

httpnneurologyorgcontent945e439fullincluding high resolution figures can be found at

References httpnneurologyorgcontent945e439fullref-list-1

This article cites 33 articles 8 of which you can access for free at

Citations httpnneurologyorgcontent945e439fullotherarticles

This article has been cited by 1 HighWire-hosted articles

Subspecialty Collections

ehttpnneurologyorgcgicollectionall_cerebrovascular_disease_strokAll Cerebrovascular diseaseStrokefollowing collection(s) This article along with others on similar topics appears in the

Permissions amp Licensing

httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online

ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology